Hi,
I'd like to send you two patches that are related to the base64 module.
I hope you find them useful.
--
The first patch fixes a bug (or at least, I think it is a bug) in the
base64_decode_alloc_ctx function. This method sometimes does not decode
the final byte in the input. Strange enough, this bug was mentioned in
the tests but not fixed.
Example (from the tests):
YWJjZGVmZ2hp decodes to abcdefghi
If you initialize a new decode-context buffer ctx and run:
base64_decode_alloc_ctx (&ctx, "YW", ...)
-> decodes nothing (that's OK)
"YW" is kept in the decode-context buffer
base64_decode_alloc_ctx (&ctx, "JjZGVmZ2", ...)
-> decodes abcdef (OK, that's "YWJjZGVm")
now the decode-context buffer contains "Z2"
then, finally:
base64_decode_alloc_ctx (&ctx, "hp", ...)
-> decodes gh (not OK, where's the final 'i'?)
I think this is because the base64_decode_alloc_ctx does not allocate
enough space in the output buffer, because it doesn't look at what is
already in the decode-context buffer. I found that if I add one extra
byte to the buffer size it works. (See patch 1.)
--
The reason I found this bug was that I was working on a base32 module. I
need a base32 encoding/decoding module, but Gnulib doesn't have one. In
patch 2, I included a new module based on the base64 code that is
already in Gnulib.
Kind regards,
Gijs van Tulder
>From d0d1a5a7a3f00e4ba418545fe9e151fb1cb357ed Mon Sep 17 00:00:00 2001
From: Gijs van Tulder <gvtul...@gmail.com>
Date: Wed, 10 Aug 2011 15:13:12 +0200
Subject: [PATCH 1/2] base64_decode_alloc_ctx would sometimes miss the final byte.
If you called base64_decode_alloc_ctx with a partial byte left over
from a previous call, the last character would not be decoded
because the output buffer was too small. (This bug was noted in
the test, stangely, but not fixed.)
---
ChangeLog | 7 +++++++
lib/base64.c | 2 +-
tests/test-base64.c | 5 ++---
3 files changed, 10 insertions(+), 4 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index 701f07b..2d165ae 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,10 @@
+2011-08-10 Gijs van Tulder <gvtul...@gmail.com>
+
+ base64: Fix bug in base64_decode_alloc_ctx, which sometimes would
+ not decode the final character.
+ * lib/base64.c: Fixed bug.
+ * lib/base64.h: Updated test.
+
2011-08-09 Bruno Haible <br...@clisp.org>
More tests for 'fseeko'.
diff --git a/lib/base64.c b/lib/base64.c
index 99fcc57..96d2cfd 100644
--- a/lib/base64.c
+++ b/lib/base64.c
@@ -555,7 +555,7 @@ base64_decode_alloc_ctx (struct base64_decode_context *ctx,
The exact size is 3 * inlen / 4, minus 1 if the input ends
with "=" and minus another 1 if the input ends with "==".
Dividing before multiplying avoids the possibility of overflow. */
- size_t needlen = 3 * (inlen / 4) + 2;
+ size_t needlen = 3 * (inlen / 4) + 3;
*out = malloc (needlen);
if (!*out)
diff --git a/tests/test-base64.c b/tests/test-base64.c
index c7cad2f..b1979b4 100644
--- a/tests/test-base64.c
+++ b/tests/test-base64.c
@@ -184,9 +184,8 @@ main (void)
ok = base64_decode_alloc_ctx (&ctx, "hp", 2, &p, &len);
ASSERT (ok);
- ASSERT (len == 2);
- /* Actually this looks buggy. Shouldn't output be 'ghi'? */
- ASSERT (memcmp (p, "gh", len) == 0);
+ ASSERT (len == 3);
+ ASSERT (memcmp (p, "ghi", len) == 0);
ok = base64_decode_alloc_ctx (&ctx, "", 0, &p, &len);
ASSERT (ok);
}
--
1.7.4.1
>From 7c2457dd2ace36b570b25ed5807d6d4a01215fad Mon Sep 17 00:00:00 2001
From: Gijs van Tulder <gvtul...@gmail.com>
Date: Wed, 10 Aug 2011 15:24:32 +0200
Subject: [PATCH 2/2] New module: base32.
---
ChangeLog | 10 +
lib/base32.c | 588 ++++++++++++++++++++++++++++++++++++++++++++++++++
lib/base32.h | 61 ++++++
m4/base32.m4 | 16 ++
modules/base32 | 26 +++
modules/base32-tests | 11 +
tests/test-base32.c | 252 +++++++++++++++++++++
7 files changed, 964 insertions(+), 0 deletions(-)
create mode 100644 lib/base32.c
create mode 100644 lib/base32.h
create mode 100644 m4/base32.m4
create mode 100644 modules/base32
create mode 100644 modules/base32-tests
create mode 100644 tests/test-base32.c
diff --git a/ChangeLog b/ChangeLog
index 2d165ae..f5b6761 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,13 @@
+2011-08-10 Gijs van Tulder <gvtul...@gmail.com>
+
+ base32: New module, based on the code for base64.
+ * lib/base32.c: New file.
+ * lib/base32.h: New file.
+ * m4/base32.m4: New file.
+ * modules/base32: New file.
+ * modules/base32-tests: New file.
+ * tests/test-base32.c: New file.
+
2011-08-10 Gijs van Tulder <gvtul...@gmail.com>
base64: Fix bug in base64_decode_alloc_ctx, which sometimes would
diff --git a/lib/base32.c b/lib/base32.c
new file mode 100644
index 0000000..351a362
--- /dev/null
+++ b/lib/base32.c
@@ -0,0 +1,588 @@
+/* base32.c -- Encode binary data using printable characters.
+ Copyright (C) 1999-2001, 2004-2006, 2009-2011 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+/* Adapted from Simon Josefsson's base64 code by Gijs van Tulder.
+ *
+ * See also RFC 3548 <http://www.ietf.org/rfc/rfc3548.txt>.
+ *
+ * Be careful with error checking. Here is how you would typically
+ * use these functions:
+ *
+ * bool ok = base32_decode_alloc (in, inlen, &out, &outlen);
+ * if (!ok)
+ * FAIL: input was not valid base32
+ * if (out == NULL)
+ * FAIL: memory allocation error
+ * OK: data in OUT/OUTLEN
+ *
+ * size_t outlen = base32_encode_alloc (in, inlen, &out);
+ * if (out == NULL && outlen == 0 && inlen != 0)
+ * FAIL: input too long
+ * if (out == NULL)
+ * FAIL: memory allocation error
+ * OK: data in OUT/OUTLEN.
+ *
+ */
+
+#include <config.h>
+
+/* Get prototype. */
+#include "base32.h"
+
+/* Get malloc. */
+#include <stdlib.h>
+
+/* Get UCHAR_MAX. */
+#include <limits.h>
+
+#include <string.h>
+
+/* C89 compliant way to cast 'char' to 'unsigned char'. */
+static inline unsigned char
+to_uchar (char ch)
+{
+ return ch;
+}
+
+/* Base32 encode IN array of size INLEN into OUT array of size OUTLEN.
+ If OUTLEN is less than BASE32_LENGTH(INLEN), write as many bytes as
+ possible. If OUTLEN is larger than BASE32_LENGTH(INLEN), also zero
+ terminate the output buffer. */
+void
+base32_encode (const char *restrict in, size_t inlen,
+ char *restrict out, size_t outlen)
+{
+ static const char b32str[32] =
+ "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567";
+
+ while (inlen && outlen)
+ {
+ *out++ = b32str[(to_uchar (in[0]) >> 3) & 0x1f];
+ if (!--outlen)
+ break;
+ *out++ = b32str[((to_uchar (in[0]) << 2)
+ + (--inlen ? to_uchar (in[1]) >> 6 : 0))
+ & 0x1f];
+ if (!--outlen)
+ break;
+ *out++ =
+ (inlen
+ ? b32str[(to_uchar (in[1]) >> 1) & 0x1f]
+ : '=');
+ if (!--outlen)
+ break;
+ *out++ =
+ (inlen
+ ? b32str[((to_uchar (in[1]) << 4)
+ + (--inlen ? to_uchar (in[2]) >> 4 : 0))
+ & 0x1f]
+ : '=');
+ if (!--outlen)
+ break;
+ *out++ =
+ (inlen
+ ? b32str[((to_uchar (in[2]) << 1)
+ + (--inlen ? to_uchar (in[3]) >> 7 : 0))
+ & 0x1f]
+ : '=');
+ if (!--outlen)
+ break;
+ *out++ =
+ (inlen
+ ? b32str[(to_uchar (in[3]) >> 2) & 0x1f]
+ : '=');
+ if (!--outlen)
+ break;
+ *out++ =
+ (inlen
+ ? b32str[((to_uchar (in[3]) << 3)
+ + (--inlen ? to_uchar (in[4]) >> 5 : 0))
+ & 0x1f]
+ : '=');
+ if (!--outlen)
+ break;
+ *out++ = inlen ? b32str[to_uchar (in[4]) & 0x1f] : '=';
+ if (!--outlen)
+ break;
+ if (inlen)
+ inlen--;
+ if (inlen)
+ in += 5;
+ }
+
+ if (outlen)
+ *out = '\0';
+}
+
+/* Allocate a buffer and store zero terminated base32 encoded data
+ from array IN of size INLEN, returning BASE32_LENGTH(INLEN), i.e.,
+ the length of the encoded data, excluding the terminating zero. On
+ return, the OUT variable will hold a pointer to newly allocated
+ memory that must be deallocated by the caller. If output string
+ length would overflow, 0 is returned and OUT is set to NULL. If
+ memory allocation failed, OUT is set to NULL, and the return value
+ indicates length of the requested memory block, i.e.,
+ BASE32_LENGTH(inlen) + 1. */
+size_t
+base32_encode_alloc (const char *in, size_t inlen, char **out)
+{
+ size_t outlen = 1 + BASE32_LENGTH (inlen);
+
+ /* Check for overflow in outlen computation.
+ *
+ * If there is no overflow, outlen >= inlen.
+ *
+ * TODO Is this a sufficient check? (See the notes in base64.c.)
+ */
+ if (inlen > outlen)
+ {
+ *out = NULL;
+ return 0;
+ }
+
+ *out = malloc (outlen);
+ if (!*out)
+ return outlen;
+
+ base32_encode (in, inlen, *out, outlen);
+
+ return outlen - 1;
+}
+
+/* With this approach this file works independent of the charset used
+ (think EBCDIC). However, it does assume that the characters in the
+ Base32 alphabet (A-Z2-7) are encoded in 0..255. POSIX
+ 1003.1-2001 require that char and unsigned char are 8-bit
+ quantities, though, taking care of that problem. But this may be a
+ potential problem on non-POSIX C99 platforms.
+
+ IBM C V6 for AIX mishandles "#define B32(x) ...'x'...", so use "_"
+ as the formal parameter rather than "x". */
+#define B32(_) \
+ ((_) == 'A' ? 0 \
+ : (_) == 'B' ? 1 \
+ : (_) == 'C' ? 2 \
+ : (_) == 'D' ? 3 \
+ : (_) == 'E' ? 4 \
+ : (_) == 'F' ? 5 \
+ : (_) == 'G' ? 6 \
+ : (_) == 'H' ? 7 \
+ : (_) == 'I' ? 8 \
+ : (_) == 'J' ? 9 \
+ : (_) == 'K' ? 10 \
+ : (_) == 'L' ? 11 \
+ : (_) == 'M' ? 12 \
+ : (_) == 'N' ? 13 \
+ : (_) == 'O' ? 14 \
+ : (_) == 'P' ? 15 \
+ : (_) == 'Q' ? 16 \
+ : (_) == 'R' ? 17 \
+ : (_) == 'S' ? 18 \
+ : (_) == 'T' ? 19 \
+ : (_) == 'U' ? 20 \
+ : (_) == 'V' ? 21 \
+ : (_) == 'W' ? 22 \
+ : (_) == 'X' ? 23 \
+ : (_) == 'Y' ? 24 \
+ : (_) == 'Z' ? 25 \
+ : (_) == '2' ? 26 \
+ : (_) == '3' ? 27 \
+ : (_) == '4' ? 28 \
+ : (_) == '5' ? 29 \
+ : (_) == '6' ? 30 \
+ : (_) == '7' ? 31 \
+ : -1)
+
+static const signed char b32[0x100] = {
+ B32 (0), B32 (1), B32 (2), B32 (3),
+ B32 (4), B32 (5), B32 (6), B32 (7),
+ B32 (8), B32 (9), B32 (10), B32 (11),
+ B32 (12), B32 (13), B32 (14), B32 (15),
+ B32 (16), B32 (17), B32 (18), B32 (19),
+ B32 (20), B32 (21), B32 (22), B32 (23),
+ B32 (24), B32 (25), B32 (26), B32 (27),
+ B32 (28), B32 (29), B32 (30), B32 (31),
+ B32 (32), B32 (33), B32 (34), B32 (35),
+ B32 (36), B32 (37), B32 (38), B32 (39),
+ B32 (40), B32 (41), B32 (42), B32 (43),
+ B32 (44), B32 (45), B32 (46), B32 (47),
+ B32 (48), B32 (49), B32 (50), B32 (51),
+ B32 (52), B32 (53), B32 (54), B32 (55),
+ B32 (56), B32 (57), B32 (58), B32 (59),
+ B32 (60), B32 (61), B32 (62), B32 (63),
+ B32 (32), B32 (65), B32 (66), B32 (67),
+ B32 (68), B32 (69), B32 (70), B32 (71),
+ B32 (72), B32 (73), B32 (74), B32 (75),
+ B32 (76), B32 (77), B32 (78), B32 (79),
+ B32 (80), B32 (81), B32 (82), B32 (83),
+ B32 (84), B32 (85), B32 (86), B32 (87),
+ B32 (88), B32 (89), B32 (90), B32 (91),
+ B32 (92), B32 (93), B32 (94), B32 (95),
+ B32 (96), B32 (97), B32 (98), B32 (99),
+ B32 (100), B32 (101), B32 (102), B32 (103),
+ B32 (104), B32 (105), B32 (106), B32 (107),
+ B32 (108), B32 (109), B32 (110), B32 (111),
+ B32 (112), B32 (113), B32 (114), B32 (115),
+ B32 (116), B32 (117), B32 (118), B32 (119),
+ B32 (120), B32 (121), B32 (122), B32 (123),
+ B32 (124), B32 (125), B32 (126), B32 (127),
+ B32 (128), B32 (129), B32 (130), B32 (131),
+ B32 (132), B32 (133), B32 (134), B32 (135),
+ B32 (136), B32 (137), B32 (138), B32 (139),
+ B32 (140), B32 (141), B32 (142), B32 (143),
+ B32 (144), B32 (145), B32 (146), B32 (147),
+ B32 (148), B32 (149), B32 (150), B32 (151),
+ B32 (152), B32 (153), B32 (154), B32 (155),
+ B32 (156), B32 (157), B32 (158), B32 (159),
+ B32 (160), B32 (161), B32 (162), B32 (163),
+ B32 (132), B32 (165), B32 (166), B32 (167),
+ B32 (168), B32 (169), B32 (170), B32 (171),
+ B32 (172), B32 (173), B32 (174), B32 (175),
+ B32 (176), B32 (177), B32 (178), B32 (179),
+ B32 (180), B32 (181), B32 (182), B32 (183),
+ B32 (184), B32 (185), B32 (186), B32 (187),
+ B32 (188), B32 (189), B32 (190), B32 (191),
+ B32 (192), B32 (193), B32 (194), B32 (195),
+ B32 (196), B32 (197), B32 (198), B32 (199),
+ B32 (200), B32 (201), B32 (202), B32 (203),
+ B32 (204), B32 (205), B32 (206), B32 (207),
+ B32 (208), B32 (209), B32 (210), B32 (211),
+ B32 (212), B32 (213), B32 (214), B32 (215),
+ B32 (216), B32 (217), B32 (218), B32 (219),
+ B32 (220), B32 (221), B32 (222), B32 (223),
+ B32 (224), B32 (225), B32 (226), B32 (227),
+ B32 (228), B32 (229), B32 (230), B32 (231),
+ B32 (232), B32 (233), B32 (234), B32 (235),
+ B32 (236), B32 (237), B32 (238), B32 (239),
+ B32 (240), B32 (241), B32 (242), B32 (243),
+ B32 (244), B32 (245), B32 (246), B32 (247),
+ B32 (248), B32 (249), B32 (250), B32 (251),
+ B32 (252), B32 (253), B32 (254), B32 (255)
+};
+
+#if UCHAR_MAX == 255
+# define uchar_in_range(c) true
+#else
+# define uchar_in_range(c) ((c) <= 255)
+#endif
+
+/* Return true if CH is a character from the Base32 alphabet, and
+ false otherwise. Note that '=' is padding and not considered to be
+ part of the alphabet. */
+bool
+isbase32 (char ch)
+{
+ return uchar_in_range (to_uchar (ch)) && 0 <= b32[to_uchar (ch)];
+}
+
+/* Initialize decode-context buffer, CTX. */
+void
+base32_decode_ctx_init (struct base32_decode_context *ctx)
+{
+ ctx->i = 0;
+}
+
+/* If CTX->i is 0 or 8, there are eight or more bytes in [*IN..IN_END), and
+ none of those eight is a newline, then return *IN. Otherwise, copy up to
+ 4 - CTX->i non-newline bytes from that range into CTX->buf, starting at
+ index CTX->i and setting CTX->i to reflect the number of bytes copied,
+ and return CTX->buf. In either case, advance *IN to point to the byte
+ after the last one processed, and set *N_NON_NEWLINE to the number of
+ verified non-newline bytes accessible through the returned pointer. */
+static inline char *
+get_8 (struct base32_decode_context *ctx,
+ char const *restrict *in, char const *restrict in_end,
+ size_t *n_non_newline)
+{
+ if (ctx->i == 8)
+ ctx->i = 0;
+
+ if (ctx->i == 0)
+ {
+ char const *t = *in;
+ if (8 <= in_end - *in && memchr (t, '\n', 8) == NULL)
+ {
+ /* This is the common case: no newline. */
+ *in += 8;
+ *n_non_newline = 8;
+ return (char *) t;
+ }
+ }
+
+ {
+ /* Copy non-newline bytes into BUF. */
+ char const *p = *in;
+ while (p < in_end)
+ {
+ char c = *p++;
+ if (c != '\n')
+ {
+ ctx->buf[ctx->i++] = c;
+ if (ctx->i == 8)
+ break;
+ }
+ }
+
+ *in = p;
+ *n_non_newline = ctx->i;
+ return ctx->buf;
+ }
+}
+
+#define return_false \
+ do \
+ { \
+ *outp = out; \
+ return false; \
+ } \
+ while (false)
+
+/* Decode eight bytes of base32-encoded data, IN, of length INLEN
+ into the output buffer, *OUT, of size *OUTLEN bytes. Return true if
+ decoding is successful, false otherwise. If *OUTLEN is too small,
+ as many bytes as possible are written to *OUT. On return, advance
+ *OUT to point to the byte after the last one written, and decrement
+ *OUTLEN to reflect the number of bytes remaining in *OUT. */
+static inline bool
+decode_8 (char const *restrict in, size_t inlen,
+ char *restrict *outp, size_t *outleft)
+{
+ char *out = *outp;
+ if (inlen < 8)
+ return false;
+
+ if (!isbase32 (in[0]) || !isbase32 (in[1]) )
+ return false;
+
+ if (*outleft)
+ {
+ *out++ = ((b32[to_uchar (in[0])] << 3)
+ | (b32[to_uchar (in[1])] >> 2));
+ --*outleft;
+ }
+
+ if (in[2] == '=')
+ {
+ if (in[3] != '=' || in[4] != '=' || in[5] != '='
+ || in[6] != '=' || in[7] != '=')
+ return_false;
+ }
+ else
+ {
+ if (!isbase32 (in[2]) || !isbase32 (in[3]))
+ return_false;
+
+ if (*outleft)
+ {
+ *out++ = ((b32[to_uchar (in[1])] << 6)
+ | (b32[to_uchar (in[2])] << 1)
+ | (b32[to_uchar (in[3])] >> 4));
+ --*outleft;
+ }
+
+ if (in[4] == '=')
+ {
+ if (in[5] != '=' || in[6] != '=' || in[7] != '=')
+ return_false;
+ }
+ else
+ {
+ if (!isbase32 (in[4]))
+ return_false;
+
+ if (*outleft)
+ {
+ *out++ = ((b32[to_uchar (in[3])] << 4)
+ | (b32[to_uchar (in[4])] >> 1));
+ --*outleft;
+ }
+
+ if (in[5] == '=')
+ {
+ if (in[6] != '=' || in[7] != '=')
+ return_false;
+ }
+ else
+ {
+ if (!isbase32 (in[5]) || !isbase32 (in[6]))
+ return_false;
+
+ if (*outleft)
+ {
+ *out++ = ((b32[to_uchar (in[4])] << 7)
+ | (b32[to_uchar (in[5])] << 2)
+ | (b32[to_uchar (in[6])] >> 3));
+ --*outleft;
+ }
+
+ if (in[7] != '=')
+ {
+ if (!isbase32 (in[7]))
+ return_false;
+
+ if (*outleft)
+ {
+ *out++ = ((b32[to_uchar (in[6])] << 5)
+ | (b32[to_uchar (in[7])]));
+ --*outleft;
+ }
+ }
+ }
+ }
+ }
+
+ *outp = out;
+ return true;
+}
+
+/* Decode base32-encoded input array IN of length INLEN to output array
+ OUT that can hold *OUTLEN bytes. The input data may be interspersed
+ with newlines. Return true if decoding was successful, i.e. if the
+ input was valid base32 data, false otherwise. If *OUTLEN is too
+ small, as many bytes as possible will be written to OUT. On return,
+ *OUTLEN holds the length of decoded bytes in OUT. Note that as soon
+ as any non-alphabet, non-newline character is encountered, decoding
+ is stopped and false is returned. If INLEN is zero, then process
+ only whatever data is stored in CTX.
+
+ Initially, CTX must have been initialized via base32_decode_ctx_init.
+ Subsequent calls to this function must reuse whatever state is recorded
+ in that buffer. It is necessary for when a octuple of base32 input
+ bytes spans two input buffers.
+
+ If CTX is NULL then newlines are treated as garbage and the input
+ buffer is processed as a unit. */
+
+bool
+base32_decode_ctx (struct base32_decode_context *ctx,
+ const char *restrict in, size_t inlen,
+ char *restrict out, size_t *outlen)
+{
+ size_t outleft = *outlen;
+ bool ignore_newlines = ctx != NULL;
+ bool flush_ctx = false;
+ unsigned int ctx_i = 0;
+
+ if (ignore_newlines)
+ {
+ ctx_i = ctx->i;
+ flush_ctx = inlen == 0;
+ }
+
+
+ while (true)
+ {
+ size_t outleft_save = outleft;
+ if (ctx_i == 0 && !flush_ctx)
+ {
+ while (true)
+ {
+ /* Save a copy of outleft, in case we need to re-parse this
+ block of four bytes. */
+ outleft_save = outleft;
+ if (!decode_8 (in, inlen, &out, &outleft))
+ break;
+
+ in += 8;
+ inlen -= 8;
+ }
+ }
+
+ if (inlen == 0 && !flush_ctx)
+ break;
+
+ /* Handle the common case of 72-byte wrapped lines.
+ This also handles any other multiple-of-8-byte wrapping. */
+ if (inlen && *in == '\n' && ignore_newlines)
+ {
+ ++in;
+ --inlen;
+ continue;
+ }
+
+ /* Restore OUT and OUTLEFT. */
+ out -= outleft_save - outleft;
+ outleft = outleft_save;
+
+ {
+ char const *in_end = in + inlen;
+ char const *non_nl;
+
+ if (ignore_newlines)
+ non_nl = get_8 (ctx, &in, in_end, &inlen);
+ else
+ non_nl = in; /* Might have nl in this case. */
+
+ /* If the input is empty or consists solely of newlines (0 non-newlines),
+ then we're done. Likewise if there are fewer than 8 bytes when not
+ flushing context and not treating newlines as garbage. */
+ if (inlen == 0 || (inlen < 8 && !flush_ctx && ignore_newlines))
+ {
+ inlen = 0;
+ break;
+ }
+ if (!decode_8 (non_nl, inlen, &out, &outleft))
+ break;
+
+ inlen = in_end - in;
+ }
+ }
+
+ *outlen -= outleft;
+
+ return inlen == 0;
+}
+
+/* Allocate an output buffer in *OUT, and decode the base32 encoded
+ data stored in IN of size INLEN to the *OUT buffer. On return, the
+ size of the decoded data is stored in *OUTLEN. OUTLEN may be NULL,
+ if the caller is not interested in the decoded length. *OUT may be
+ NULL to indicate an out of memory error, in which case *OUTLEN
+ contains the size of the memory block needed. The function returns
+ true on successful decoding and memory allocation errors. (Use the
+ *OUT and *OUTLEN parameters to differentiate between successful
+ decoding and memory error.) The function returns false if the
+ input was invalid, in which case *OUT is NULL and *OUTLEN is
+ undefined. */
+bool
+base32_decode_alloc_ctx (struct base32_decode_context *ctx,
+ const char *in, size_t inlen, char **out,
+ size_t *outlen)
+{
+ /* This may allocate a few bytes too many, depending on input,
+ but it's not worth the extra CPU time to compute the exact size.
+ The exact size is 5 * inlen / 8, minus one or more bytes if the
+ input is padded with one or more "=".
+ Dividing before multiplying avoids the possibility of overflow. */
+ size_t needlen = 5 * (inlen / 8) + 5;
+
+ *out = malloc (needlen);
+ if (!*out)
+ return true;
+
+ if (!base32_decode_ctx (ctx, in, inlen, *out, &needlen))
+ {
+ free (*out);
+ *out = NULL;
+ return false;
+ }
+
+ if (outlen)
+ *outlen = needlen;
+
+ return true;
+}
diff --git a/lib/base32.h b/lib/base32.h
new file mode 100644
index 0000000..8768686
--- /dev/null
+++ b/lib/base32.h
@@ -0,0 +1,61 @@
+/* base32.h -- Encode binary data using printable characters.
+ Copyright (C) 2004-2006, 2009-2011 Free Software Foundation, Inc.
+ Adapted from Simon Josefsson's base64 code by Gijs van Tulder.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#ifndef BASE32_H
+# define BASE32_H
+
+/* Get size_t. */
+# include <stddef.h>
+
+/* Get bool. */
+# include <stdbool.h>
+
+/* This uses that the expression (n+(k-1))/k means the smallest
+ integer >= n/k, i.e., the ceiling of n/k. */
+# define BASE32_LENGTH(inlen) ((((inlen) + 4) / 5) * 8)
+
+struct base32_decode_context
+{
+ unsigned int i;
+ char buf[8];
+};
+
+extern bool isbase32 (char ch);
+
+extern void base32_encode (const char *restrict in, size_t inlen,
+ char *restrict out, size_t outlen);
+
+extern size_t base32_encode_alloc (const char *in, size_t inlen, char **out);
+
+extern void base32_decode_ctx_init (struct base32_decode_context *ctx);
+
+extern bool base32_decode_ctx (struct base32_decode_context *ctx,
+ const char *restrict in, size_t inlen,
+ char *restrict out, size_t *outlen);
+
+extern bool base32_decode_alloc_ctx (struct base32_decode_context *ctx,
+ const char *in, size_t inlen,
+ char **out, size_t *outlen);
+
+#define base32_decode(in, inlen, out, outlen) \
+ base32_decode_ctx (NULL, in, inlen, out, outlen)
+
+#define base32_decode_alloc(in, inlen, out, outlen) \
+ base32_decode_alloc_ctx (NULL, in, inlen, out, outlen)
+
+#endif /* BASE32_H */
diff --git a/m4/base32.m4 b/m4/base32.m4
new file mode 100644
index 0000000..efa0ef7
--- /dev/null
+++ b/m4/base32.m4
@@ -0,0 +1,16 @@
+# base32.m4 serial 3
+dnl Copyright (C) 2004, 2006, 2009-2011 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+AC_DEFUN([gl_FUNC_BASE32],
+[
+ gl_PREREQ_BASE32
+])
+
+# Prerequisites of lib/base32.c.
+AC_DEFUN([gl_PREREQ_BASE32], [
+ AC_REQUIRE([AC_C_INLINE])
+ AC_REQUIRE([AC_C_RESTRICT])
+])
diff --git a/modules/base32 b/modules/base32
new file mode 100644
index 0000000..20f38cf
--- /dev/null
+++ b/modules/base32
@@ -0,0 +1,26 @@
+Description:
+Encode binary data using printable characters (base32).
+
+Files:
+lib/base32.h
+lib/base32.c
+m4/base32.m4
+
+Depends-on:
+stdbool
+memchr
+
+configure.ac:
+gl_FUNC_BASE32
+
+Makefile.am:
+lib_SOURCES += base32.h base32.c
+
+Include:
+"base32.h"
+
+License:
+LGPLv2+
+
+Maintainer:
+TODO
diff --git a/modules/base32-tests b/modules/base32-tests
new file mode 100644
index 0000000..1422395
--- /dev/null
+++ b/modules/base32-tests
@@ -0,0 +1,11 @@
+Files:
+tests/test-base32.c
+tests/macros.h
+
+Depends-on:
+
+configure.ac:
+
+Makefile.am:
+TESTS += test-base32
+check_PROGRAMS += test-base32
diff --git a/tests/test-base32.c b/tests/test-base32.c
new file mode 100644
index 0000000..e7cc5cc
--- /dev/null
+++ b/tests/test-base32.c
@@ -0,0 +1,252 @@
+/* Self tests for base32.
+ Copyright (C) 2004, 2008-2011 Free Software Foundation, Inc.
+ Based on the tests for base64 written by Simon Josefsson.
+ Adapted for base32 by Gijs van Tulder.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#include "base32.h"
+
+#include <stddef.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdint.h>
+
+#include "macros.h"
+
+int
+main (void)
+{
+ const char *in = "abcdefghijklmnop";
+ const char *b32in = "MFRGGZDFMZTWQ2LKNNWG23TPOA======";
+ char out[255];
+ size_t len;
+ bool ok;
+ char *p;
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 0, out, 0);
+ ASSERT (out[0] == '\x42');
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 1, out, 10);
+ ASSERT (memcmp (out, "ME======", 1) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 1, out, 2);
+ ASSERT (memcmp (out, "ME======", 2) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 1, out, 3);
+ ASSERT (memcmp (out, "ME======", 3) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 1, out, 4);
+ ASSERT (memcmp (out, "ME======", 4) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 1, out, 8);
+ ASSERT (memcmp (out, "ME======", 8) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 2, out, 8);
+ ASSERT (memcmp (out, "MFRA====", 8) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 3, out, 8);
+ ASSERT (memcmp (out, "MFRGG===", 8) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 4, out, 8);
+ ASSERT (memcmp (out, "MFRGGZA=", 8) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 5, out, 8);
+ ASSERT (memcmp (out, "MFRGGZDF", 8) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 6, out, 16);
+ ASSERT (memcmp (out, "MFRGGZDFMY======", 16) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ base32_encode (in, 6, out, 100);
+ ASSERT (memcmp (out, "MFRGGZDFMY======", 16) == 0);
+
+ /* Decode. */
+
+ memset (out, 0x42, sizeof (out));
+ len = 0;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 1;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 1);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 1) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 2;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 2);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 2) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 3;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 3);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 3) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 4;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 4);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 4) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 5;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 5);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 5) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 6;
+ ok = base32_decode (b32in, 8, out, &len);
+ ASSERT (ok);
+ ASSERT (len == 5);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 5) == 0);
+
+ memset (out, 0x42, sizeof (out));
+ len = 100;
+ ok = base32_decode (b32in, strlen (b32in), out, &len);
+ ASSERT (ok);
+ ASSERT (len == 16);
+ ASSERT (memcmp (out, "abcdefghijklmnop", 16) == 0);
+
+ /* Allocating encode */
+
+ len = base32_encode_alloc (in, strlen (in), &p);
+ ASSERT (len == 32);
+ ASSERT (strcmp (p, "MFRGGZDFMZTWQ2LKNNWG23TPOA======") == 0);
+ free (p);
+
+ len = base32_encode_alloc (in, SIZE_MAX - 5, &p);
+ ASSERT (len == 0);
+
+ /* Decode context function */
+ {
+ struct base32_decode_context ctx;
+
+ base32_decode_ctx_init (&ctx);
+
+ len = sizeof (out);
+ ok = base32_decode_ctx (&ctx, b32in, strlen (b32in), out, &len);
+ ASSERT (ok);
+ ASSERT (len == 16);
+ ASSERT (memcmp (out, "abcdefghijklmnop", len) == 0);
+ }
+
+ /* Allocating decode context function */
+
+ ok = base32_decode_alloc_ctx (NULL, b32in, strlen (b32in), &p, &len);
+ ASSERT (ok);
+ ASSERT (len == 16);
+ ASSERT (memcmp (out, "abcdefghijklmnop", len) == 0);
+
+ {
+ struct base32_decode_context ctx;
+ const char *newlineb32 = "MFRG\nGZDFMZTWQ2LKNNW\nG23TPOA======";
+
+ base32_decode_ctx_init (&ctx);
+
+ ok = base32_decode_alloc_ctx (&ctx, newlineb32, strlen (newlineb32), &p, &len);
+ ASSERT (ok);
+ ASSERT (len == strlen (in));
+ ASSERT (memcmp (p, in, len) == 0);
+ }
+
+ {
+ struct base32_decode_context ctx;
+ base32_decode_ctx_init (&ctx);
+
+ ok = base32_decode_alloc_ctx (&ctx, "MFRGGZDFM\nZTWQ2LK", 17, &p, &len);
+ ASSERT (ok);
+ ASSERT (len == 10);
+ ASSERT (memcmp (p, "abcdefghij", len) == 0);
+
+ base32_decode_ctx_init (&ctx);
+
+ ok = base32_decode_alloc_ctx (&ctx, "MF\n", 3, &p, &len);
+ ASSERT (ok);
+ ASSERT (len == 0);
+
+ ok = base32_decode_alloc_ctx (&ctx, "RGGZDFMZ", 8, &p, &len);
+ ASSERT (ok);
+ ASSERT (len == 5);
+ ASSERT (memcmp (p, "abcde", len) == 0);
+
+ ok = base32_decode_alloc_ctx (&ctx, "TWQ2LK", 6, &p, &len);
+ ASSERT (ok);
+ ASSERT (len == 5);
+ ASSERT (memcmp (p, "fghij", len) == 0);
+ ok = base32_decode_alloc_ctx (&ctx, "", 0, &p, &len);
+ ASSERT (ok);
+ }
+
+ {
+ struct base32_decode_context ctx;
+ const char *newlineb32 = "\n\n\n\n\n";
+
+ base32_decode_ctx_init (&ctx);
+
+ ok = base32_decode_alloc_ctx (&ctx, newlineb32, strlen (newlineb32), &p, &len);
+ ASSERT (ok);
+ ASSERT (len == 0);
+ }
+
+ ok = base32_decode_alloc_ctx (NULL, " ! ", 3, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "ABC\nDEF", 7, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "AA", 2, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "AA=", 3, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "AABBAAxx", 8, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "AABBAA=X", 8, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "AABBAA=X", 8, &p, &len);
+ ASSERT (!ok);
+
+ ok = base32_decode_alloc_ctx (NULL, "AABBAA=A", 8, &p, &len);
+ ASSERT (!ok);
+
+ return 0;
+}
--
1.7.4.1
